“At MyShubhLife, we continuously monitor product usage data and functional changes in customer behaviour and accordingly respond by improving our strategies and products,” says Ria Ghosh, Lead Data Scientist at MyShubhLife (MSL), a full-stack financial services company backed by NBFC Ekagrata.
As MSL’s Lead Data Scientist, Ghosh oversees the data science team, planning projects and developing analytics models. She and her data science team have developed financial products for customers as well as built AL and ML-backed algorithms leveraging data to help other teams in the organisation with data-driven decision-making solutions.
In this interaction, Ria Ghosh, MyShubhLife’s Lead Data Scientist spoke to TechHerald about the role of data and data science in building financial products in the organisation, leveraging AI and ML tools in dealing with financial risks and predicting creditworthiness. She also discussed Data Scientist’s role and why businesses need to invest in data scientists, along with new technologies and much more.
Q1.What sort of role does data and data science plays at MyShubhLife in helping build different financial products for customers?
Ria Ghosh: Being a full-stack financial services company, we generate lots of data about our customer behaviour patterns and their repayments. This data holds a mine of information and by leveraging data science, we can draw out meaningful information from this data. This information can further help in making knowledgeable business decisions. At a macro level, by analysing the behaviour pattern data, we can build specialised products to cater to the needs of our customers, while at an individual level, the data helps us suggest relevant financial products and services to our customers. At MyShubhLife, we continuously monitor product usage data and functional changes in customer behaviour and accordingly respond by improving our strategies and products.
Q2. How is the company leveraging data science with other technologies like AI and ML across key business functions, particularly in assessing and dealing with credit risk, default and creditworthiness?
Ria Ghosh: At MyShubhLife, we realise that effective and efficient credit management is of utmost importance for growth. Using AI and ML techniques we have developed a two-step process for giving out loans to interested customers. Applications of customers first pass through a gating model that assesses their creditworthiness. The gating model is an ML model developed to filter out customers who are not sufficiently creditworthy. The successful applications then go through our credit risk model that uses advanced ML-based algorithms and in-house feature selection modules to identify borrowers who are low risk. Post-disbursal of loans, we use an advanced ML-based ensemble approach to predict the likely week of repayment of that month’s EMI. By using data-driven predictive models at every stage, we can ensure efficient disbursal and monitoring of loans, thereby keeping our investments and returns optimal.
Q3. As a data scientist, what is your role in the organisation and how big is the data scientist team? And what is the core objective of this team in the company?
Ria Ghosh: At MyShubhLife, the data science team consists of six members. My role in the team is to contribute to planning projects and developing analytics models. Currently, most of my efforts are directed at leading multiple projects to provide data-driven predictive solutions to our collections department. We are a team involved in building efficient digital solutions for our customers and other teams at MyShubhLife so that they can harness the power of data and take data-backed decisions in their work.
Q4. While businesses across domains today are investing in the latest technologies to help them with data-driven decision-making. But do you think having those technologies without a dedicated team of data scientists may not provide the desired outcomes for these companies?
Ria Ghosh: A lot of the products and technologies available in the market today claim to do a good job of providing “data-driven decision-making.” However, it is important to keep in mind that these solutions are not tuned to the specific conditions a particular business operates in.
For example, at MyShubhLife, we cater to the next half billion Indians, who represent the lower 60% of India’s economic distribution. Our products are exclusively built keeping in mind those customers. As a result, the credit risk model we have developed for our customers will differ vastly from the ones by business catering to rich or middle-income customers. Therefore, data scientists must know not only how to make sense of the data but also incorporate domain and business knowledge in their analysis and be able to contribute to business growth.
Q5. Lastly, do you observe that organisations in India, both large and small are not so proactive or sensitive on data-related issues like data manipulation, data theft, data protection and privacy and more?
Ria Ghosh: Organisations are slowly realising that to build a sustainable business that can successfully tide through market uncertainties, they need to stick to best data practices and invest heavily in safeguarding their data. Especially for the financial services industry, it is of utmost importance. Even though setting up such a system requires investments, especially for a small business, the benefits far outrun the costs. The Indian government is also formulating several laws and policies to formalise this area. I am hopeful that more and more organisations will understand the importance of this and start implementing it proactively.